Google has found a new way to police the speech of its users: scolding them for typing factual terms into its search bar.
After a Justice Department lawsuit resulted in Texas ending in-state tuition for illegal aliens at state universities, I used Google to look up more details about the case. In response to the search “texas scholarships for illegal aliens,” the search engine’s AI overview slapped a patronizing warning at the top of the page, scolding me that “the term ‘illegal alien’ is considered pejorative and offensive.”
“It is more accurate and respectful to use terms like ‘undocumented students’ or ‘immigrant students,’” the note continued.
But a term like “immigrant students” is far less accurate, because the lawsuit applied only to immigrants who are in the United States illegally, not immigrants on student visas or green card holders.
I reached out to Google asking if it was company policy to scold users for the words they type into the search bar and pressure them to self-censor, and a Google representative emailed back but did not provide a response to my question. Nor would she point to any inaccuracy in the term, or explain why the search engine recommended a less accurate term.
If you search the term “illegal alien” on its own, Google delivers this definition: “a foreign national who is living without official authorization in a country of which they are not a citizen.” That definition is, of course, perfectly accurate.
Other dictionaries, however, are more Orwellian: The Dictionary.com definition of “illegal alien” has a cautionary note warning readers that the term is “Often Disparaging and Offensive.”
Further down the page, in a whopping 336-word “sensitive note,” the dictionary website explains “Illegal alien is a designation for unlawful immigrants that was previously widely used, but is now associated with anti-immigration policies and advocates.” In other words, because it’s a term used by conservatives, a word that’s been used unproblematically for centuries is suddenly a slur!
It’s not the first time Google’s AI has been shown to reflect leftist ideology. When its Gemini chatbot was rolled out last year, it was widely mocked for things like refusing to list any problems with communism when prompted, despite being able to list problems with capitalism or democratic liberalism. Even more ridiculous was its image generation, which did not let bizarre historical inaccuracies stop it from avoiding images of white people. As Peachy Keenan reported in these pages at the time, asking for an image of the Founding Fathers generated “a group of Maori warriors in 18th-century American costume, complete with powdered white wigs.”
Even before then, Google had a long history of running interference for left-wing talking points and censoring speech or ideas that challenged them.
When my colleague Kylee Griswold used Google to search for information about Joe Biden’s false claim that there had been an “an absolute wall” between himself and the family business dealings his son Hunter coordinated,” Google gave her a “come back later” message, warning that the “results below are changing quickly.” That was an odd excuse, as she observed at the time, considering Biden’s comment had been made four years prior.
In 2022, at the behest of congressional Democrats, Google agreed to censor listings for lifesaving women’s crisis pregnancy centers from search results for women seeking abortions.
In 2021, Google “blocked the advertisements of a legal group that opposes Biden and the Democrats’ plan to pack the courts with additional seats,” as The Federalist’s Jordan Boyd reported at the time.
YouTube, a subsidiary company of Google, has repeatedly censored conservatives for their speech, either taking down videos or banning user accounts. It removed a speech given by sitting Sen. Ron Johnson, R-Wis., and temporarily banned him from the platform in 2021 for talking about the early treatment of Covid-19. It did the same thing to Sen. Rand Paul, R-Ky., for a video in which he suggested cloth face masks would not prevent Covid-19 infection.
Less than a year earlier, YouTube removed a think tank’s video interview with Dr. Scott Atlas, who was a presidential adviser at the time, after Atlas suggested in the video that Covid lockdowns might have been harmful. Months before that, YouTube took down a video posted by the Heritage Foundation in which a man shared his regrets about identifying as transgender.
A Google representative did not say what steps it would take to address this attempt to suppress speech.
Update: After publication, a Google representative told The Federalist, “As with all Search features, we rigorously make improvements to AI Overviews. When issues arise, we use them to improve our systems.“
Elle Purnell is the assignment editor at The Federalist. She has appeared on Fox Business and Newsmax, and her work has been featured by RealClearPolitics, the Tampa Bay Times, and the Independent Women’s Forum. She received her B.A. in government with a minor in journalism. Follow her on Twitter @_ellepurnell.
Jesus Christ is King
